System and method for identifying miscategorization

Abstract

A categorization analysis system is provided. The categorization analysis system includes one or more hardware processors, a memory including a first plurality of listings categorized in a first target category, and a categorization analysis engine executing on the one or more hardware processors. The categorization analysis engine is configured to determine a label for each listing including performing a search on title, select a set of training listings based on the determined labels, train a first model using the set of training listings and the determined labels, the first model being a classification model configured to classify categorization of listings, identify a suspect listing categorized in the first target category, apply the suspect listing to the first model, thereby generating a categorization result for the suspect listing, the categorization result indicating miscategorization of the suspect listing, and identify the suspect listing in the memory as miscategorized.

The invention claimed is: 
     
       1. A system for categorization analysis comprising:
 hardware processing circuitry configured to perform operations comprising:
 receiving first input assigning a first category to a listing; 
 receiving second input assigning a second category to the listing; 
 analyzing the listing using a first trained model for the first category to determine an indication of whether the listing is properly categorized with respect to the first category; 
 analyzing the listing using a second trained model for the second category to determine an indication of whether the listing is properly categorized with respect to the second category; and 
 removing the listing from the second category based on a determination by the second trained model that the listing is not properly categorized in the second category while maintaining the listing in the first category in response to the first trained model determining the listing is properly categorized in the first category. 
 
 
     
     
       2. The system of  claim 1 , wherein the first input and the second input are received via a user interface from a seller for the listing. 
     
     
       3. The system of  claim 1 , the operations further comprising outputting the listing to a seller or to a site administrator in response to the determination by the second trained model that the listing is not properly categorized in the second category. 
     
     
       4. The system of  claim 1 , the operations further comprising removing the listing from the first category based on a determination by the first trained model that the listing is not properly categorized in the first category. 
     
     
       5. The system of  claim 1 , the operations further comprising excluding the listing from search results for the first category based on a determination by the first trained model that the listing is not properly categorized in the first category. 
     
     
       6. The system of  claim 1 , the operations further comprising demoting the listing in search results for the first category based on a determination by the first trained model that the listing is not properly categorized in the first category. 
     
     
       7. The system of  claim 1 , the operations further comprising:
 identifying a predetermined number of highest ranking historical listings based on a title of the listing; 
 determining a first percentage of the predetermined number of highest ranking historical listings that are categorized in the first category; and 
 determining the listing is not properly categorized in the first category if the first percentage is below a threshold. 
 
     
     
       8. The system of  claim 7 , wherein the identification of the predetermined number of highest ranking historical listings is further based on a price of the listing. 
     
     
       9. The system of  claim 7 , the operations further comprising training the first trained model based on the predetermined number of highest ranking historical listings. 
     
     
       10. A computer-implemented method of verifying categorization of an on-line listing, comprising:
 receiving first input assigning a first category to the on-line listing; 
 receiving second input assigning a second category to the on-line listing; 
 analyzing, by hardware processing circuitry, the listing using a first trained model for the first category to determine an indication of whether the listing is properly categorized with respect to the first category; 
 analyzing the listing using a second trained model for the second category to determine an indication of whether the listing is properly categorized with respect to the second category; and 
 excluding the listing from search results for the first category based on a determination by the first trained model that the listing is not properly categorized in the first category while including the listing in search results for the second category based on a determination by the second trained model that the listing is properly categorized in the second category. 
 
     
     
       11. The method of  claim 10 , further comprising outputting the listing to a seller or to a site administrator in response to the determination by the second trained model that the listing is not properly categorized in the second category. 
     
     
       12. The method of  claim 10 , further comprising removing the listing from the second category based on a determination by the second trained model that the listing is not properly categorized in the second category. 
     
     
       13. The method of  claim 10 , further comprising demoting the listing in search results for the second category based on a determination by the second trained model that the listing is not properly categorized in the second category. 
     
     
       14. The method of  claim 10 , further comprising:
 identifying a predetermined number of highest ranking historical listings based on a title of the listing; 
 determining a first percentage of the predetermined number of highest ranking historical listings that are categorized in the first category; and 
 determining the listing is not properly categorized in the first category if the first percentage is below a threshold. 
 
     
     
       15. The method of  claim 14 , wherein the identification of the predetermined number of highest ranking historical listings is further based on a price of the listing. 
     
     
       16. A non-transitory computer readable storage medium comprising instructions that when executed cause hardware processing circuitry to perform operations comprising:
 receiving first input assigning a first category that describes a first class of merchandise to a listing; 
 receiving second input assigning a second category that describes a second class of merchandise to the listing; 
 analyzing the listing using a first trained model for the first category to determine an indication of whether the listing is properly categorized with respect to the first category; 
 analyzing the listing using a second trained model for the second category to determine an indication of whether the listing is properly categorized with respect to the second category; and 
 removing the listing from the second category based on a determination by the second trained model that the listing is not properly categorized in the second category while maintaining the listing in the first category in response to the first trained model determining the listing is properly categorized in the first category. 
 
     
     
       17. The non-transitory computer readable storage medium of  claim 16 , wherein the first input and the second input are received via a user interface from a seller for the listing. 
     
     
       18. The non-transitory computer readable storage medium of  claim 16 , the operations further comprising outputting the listing to a seller or to a site administrator in response to the determination by the second trained model that the listing is not properly categorized in the second category. 
     
     
       19. The non-transitory computer readable storage medium of  claim 16 , the operations further comprising removing the listing from the first category based on a determination by the first trained model that the listing is not properly categorized in the first category. 
     
     
       20. The non-transitory computer readable storage medium of  claim 16 , the operations further comprising excluding the listing from search results for the first category based on a determination by the first trained model that the listing is not properly categorized in the first category.